Currently, with the rapid development of the Internet, virtualization and big data processing are attracting increasingly more attention in the industry. The combination of the two technologies becomes a trendy contemporary research direction in the industry. A big data processing solution may be deployed and running in a virtual environment to greatly improve the system resource utilization rate and the configuration flexibility. However, compared with a physical environment, the virtual environment has greater uncertainty, therefore the high availability of the system becomes ever more important.
In the prior art, in order to maintain the high availability of a virtual system, a physical host machine of the virtual machine will usually be monitored to determine whether the host machine is faulty.
However, when using the above-mentioned method for monitoring the physical host machine of the virtual machine, and the host machine is faulty, all virtual machines on the physical host machine need to be manually repaired by operation and maintenance personnel, thereby wasting a relatively long time, resulting in long service suspension and bringing loss to users.